Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Holo3-122B-A10B
~79
Winner · 1/8 categoriesPhi-4
40
0/8 categoriesHolo3-122B-A10B· Phi-4
Pick Holo3-122B-A10B if you want the stronger benchmark profile. Phi-4 only becomes the better choice if you want the cheaper token bill.
Holo3-122B-A10B is clearly ahead on the aggregate, 79 to 40. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Holo3-122B-A10B's sharpest advantage is in agentic, where it averages 78.9 against 38.3. The single biggest benchmark swing on the page is OSWorld-Verified, 78.8% to 34%.
Holo3-122B-A10B is also the more expensive model on tokens at $0.40 input / $3.00 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Phi-4. That is roughly Infinityx on output cost alone. Holo3-122B-A10B gives you the larger context window at 64K, compared with 16K for Phi-4.
BenchLM keeps the benchmark table and the operator tradeoffs on the same page so a better score does not hide a materially slower, pricier, or smaller-context model.
Runtime metrics show N/A when BenchLM does not have a sourced snapshot for that exact model. The scoring rules and freshness policy are documented on the methodology page.
| Benchmark | Holo3-122B-A10B | Phi-4 |
|---|---|---|
| AgenticHolo3-122B-A10B wins | ||
| OSWorld-Verified | 78.8% | 34% |
| Terminal-Bench 2.0 | — | 44% |
| BrowseComp | — | 35% |
| Coding | ||
| HumanEval | — | 82.6% |
| SWE-bench Pro | — | 55% |
| Multimodal & Grounded | ||
| MMMU-Pro | — | 54% |
| OfficeQA Pro | — | 38% |
| Reasoning | ||
| LongBench v2 | — | 30% |
| MRCRv2 | — | 33% |
| Knowledge | ||
| MMLU | — | 84.8% |
| GPQA | — | 56.1% |
| FrontierScience | — | 52% |
| Instruction Following | ||
| Coming soon | ||
| Multilingual | ||
| MGSM | — | 80.6% |
| MMLU-ProX | — | 60% |
| Mathematics | ||
| MATH-500 | — | 94.6% |
Holo3-122B-A10B is ahead overall, 79 to 40. The biggest single separator in this matchup is OSWorld-Verified, where the scores are 78.8% and 34%.
Holo3-122B-A10B has the edge for agentic tasks in this comparison, averaging 78.9 versus 38.3. Inside this category, OSWorld-Verified is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.